search

Seokju-Cho / Volumetric-Aggregation-Transformer

Official Implementation of VAT
MIT License
153 stars 14 forks source link

rescaling the predicted keypoints coordinates #9

Closed Brahmagupta closed 1 year ago

Brahmagupta commented 1 year ago

I'm trying to visualize some test results by using /utils_training/utils.flow2kps method. and I figured your model is trying to map keypoints from target images to src images, i.e. making predictions of keypoints in src images. (if im not mistaken). but the Estimated_kps Im getting seem off and sometimes out of range (out side of the image regions). I also noticed both the source and target image in a pair are resized to a same image size. this is prbly something I missed

I wonder how i can rescale the estimated_kps so it will correctly mark out the keypoints in the original images. (Orwhat size should i rescale the original image to use the predicted keypoints)

SunghwanHong commented 1 year ago

Hi, sorry for the late reply. If you see this code : flow = F.interpolate(flow, upsample_size, mode='bilinear') (upsample_size[0] / h) We only recale it using "h". This means that we assumed h==w. So if you want to obtain keypoints that are h!=w, then you could do something like this flow[0] = F.interpolate(flow[0], upsample_size[0], mode='bilinear') (upsample_size[0] / h) flow[1] = F.interpolate(flow[1], upsample_size[1], mode='bilinear') * (upsample_size[1] / w)